Personalized medicine will require the use of discriminative tests that are predictive of disease state and individual treatment response. Major depressive disorder (MDD) is a heterogeneous illness that is the leading cause of disability in developed countries and a major cause of premature death due to suicide. Research aimed at characterizing the pathology of MDD however, has mostly focused on central mechanisms and has only occasionally been translated into hypotheses that can be tested in clinical settings using readily available peripheral blood samples from human subjects (and with modest results for neuroendocrine tests for instance). Focusing on altered transcriptome in relevant brain areas, we have now identified gene expression changes that are associated with depressed states in both the human postmortem brain and the unpredictable chronic mild stress (UCMS) rodent model of depression, and that are reversed by antidepressant (AD) treatments in mice. The presence of this conserved biosignature of MDD demonstrates that MDD is associated with persistent molecular pathologies, and that a parallel study in a more tractable animal model of the illness can support the analysis of the complex human data in identifying conserved and state-dependent disease-related changes. How can we translate these findings into a potential clinical tool? The identification of peripheral biomarkers co-varying with disease states and treatment response is a primary goal in developing a biological (in addition to symptomatic) definition of MDD. Studies showed that blood transcriptomes largely parallel central profiles and can provide valuable information on subject-specific parameters, including for neuropsychiatric disorders. Hence, based on our prior findings in the brain, we hypothesize that MDD-related blood changes will manifest as conserved gene changes, for which the UCMS rodent model will provide an independent line of validation and a means to assess confounding effects of AD treatments, together facilitating the identification of MDD- and relevant gene transcript changes as putative peripheral disease biomarkers. To optimize our probability of success, we will assess patients with MDD and co-occurring anxiety receiving citalopram treatment and psychotherapy (to augment AD compliance). Blood transcriptome changes will be assessed in patients (n=40) before and after AD treatment and in control subjects (n=20) (Aim 1) to investigate the potential of peripheral gene changes to track disease state and treatment response/non-response. Parallel studies in UCMS-exposed and AD-treated mice will help us distinguish candidate biomarkers from confounding AD effects in human subjects, and will begin assessing brain-blood conservation of changes (Aim 2). Follow-up studies will assess the sensitivity and disease specificity of the identified biomarkers, extend findings from the response to the remission timeframe, and will lead to biological predictions about disease state, risk factors and treatment responses, which will be tested in future studies in larger clinical cohorts.